ICES mar. Sei. Symp., 196: 62-67. 1993

The influence of seismic exploration with airguns on cod (Gadus morhua) behaviour and catch rates

Svein Løkkeborg and Aud Void Soldai

Løkkeborg, S., and Soldai, A. V. 1993. The influence of seismic exploration with airguns on cod (Gadus morhua) behaviour and catch rates. - ICES mar. Sei. Symp., 196: 62-67.

Analyses of catch records showed that geophysical activity with airguns significantly influenced the catch rates of cod (Gadus morhua) in long-line and trawl fisheries. Catch reductions of 55-80% were observed for long lines set within a seismic survey area, and the by-catch of cod in shrimp trawl was reduced by about 80-85%. The by- catch of cod in trawl fishery for saithe, however, was observed to increase threefold and to return to normal immediately after the seismic work ended. The predominant frequencies of airgun sound spectra match the most sensitive auditory band of cod. The reductions in catch rates are undoubtedly due to behavioural responses of cod to airgun sound. The fish probably avoided the approaching geophysical vessel by swimming away from the sound source, and the amount of fish available to any fishing gear used in this area was thereby reduced. The results from the trawl fishery for saithe were explained by the short duration of the sound emissions during this particular seismic survey.

Svein Løkkeborg and A u d Void Soldai: Institute o f Marine Research, Fish Capture Division, PO Box 1870 Nordnes, N-5024 Bergen, Norway.

Introduction

Geophysical surveys have been conducted continuously in the North Sea for about 30 years. During the last few years, seismic survey vessels have also been operating on fishing grounds in the Norwegian Sea and in the Barents Sea. Concern has been expressed by fishermen that sounds generated by acoustic survey devices affect com­mercial fishing. In particular, Norwegian fishermen using long line or trawl have reported significantly reduced catch rates caused by the operations of seismic vessels.

Fish detect and respond to sounds (Tavolga et al. , 1981; Hawkins, 1986), and airgun discharges have been reported as eliciting startle and alarm responses in rockfish (Sebastes spp.) (Pearson et a i , 1992). Changes in the depth distribution of whiting (Merlangius merlan- gus) upon discharges of a single airgun have been reported (Chapman and Hawkins, 1969), and a field study suggested changes in fish distribution along the tracklines of a seismic survey vessel (Dalen and Raknes,1985). In hook and line fishing, the catch rate of rockfish was observed to decline by 50% during sound emission from a single airgun (Skalski et al., 1992). However, the

quantitative effects of seismic survey operations on catch success in commercial long-lining and trawling have not been documented.

The present study is based on catch data obtained from long-liners and trawlers fishing in areas where seismic survey operations were being conducted. Catch rates of cod (Gadus morhua) obtained within and at various distances from a seismic survey area were com­pared. The duration of the effect of seismic operations was also investigated.

Materials and methods

Long-lining

In January 1990, a seismic survey was carried out off the coast of Finnmark (northern Norway) in an area in which several long-liners were fishing for cod. The seismic vessel used an array of four sleeve guns (chamber volume 40 in3). The survey tracklines were about 4700 m long, and the guns were discharged every 12.5 m (i.e. at intervals of about 5 s). A total of 32 tracklines were run during four periods (Table 1). Catch data were obtained from four autoliners (“Husby” ,

62

Table 1. Schedule of the seismic survey conducted off the coast of Finnmark (northern Norway) in 1990.

Start sound End soundemission emission No. of

Period Date Time Date Timesurvey

tracklines

1 19 Jan 18.32 20 Jan 04.33 72 21 Jan 16.43 22 Jan 06.38 103 28 Jan 09.14 28 Jan 12.00 34 29 Jan 03.44 29 Jan 20.27 12

“Førde” , “Frøyanes” , and “Værland”). For each fleet of long lines, the skipper normally notes the position, time of setting, and estimates the weight of the catch. These notes were related to the positions of the survey track­lines and the times of discharges of the guns.

Trawling

The effects of seismic operations on the by-catch of cod above legal size (>42 cm) in trawl fisheries for shrimp (two cases) and saithe (one case) were studied. With the permission of the shipowner, catch data were obtained from the official catch records required and stored by the Norwegian authorities. The estimated weight of the catch, and the time and start position of each haul, were noted and related to the positions and times of running of the geophysical survey tracklines. As the catch rec­ords only give the start position of each haul, and not the towing direction or stop position, it is not possible to calculate the exact distance between the survey track­lines and the trawl haul. Each trawl haul is usually between 10 and 15 nautical miles long. In these analyses, hauls starting closer than 5 nautical miles from the seismic tracklines are considered to have been influ­enced by the airgun noise. Variance analyses (General Linear Models, SAS Institute Inc., Cary, IN) were used to test differences between catch rates before, during and after seismic activity.

The following cases were analysed:

1. A geophysical survey conducted off the coast of Finnmark, northern Norway, in June 1989. The seis­mic source used was an array of 20 sleeve guns each with a chamber volume of 40 in3. Shot interval was 25 m. A shrimp trawler was fishing within and close to the survey area (bottom depth 200-300 m) several weeks before and for two days during the seismic activity. Catch data from hauls taken before and during the airgun activity were compared.

2. A survey conducted in the Barents Sea east of Bear Island in August 1991. The seismic vessel used an array of 40 airguns with a total volume of 2660 in3 and a shot interval of 25 m. A shrimp trawler, fishingwithin the Bear Island area at 200 to 300 m depth for several weeks, was trawling closer than 5 nautical miles from the geophysical vessel for three days.

Catch records from before, during, and after the seismic activity were compared.

3. A survey conducted at Storegga of the coast of Møre (western Norway) in April 1991. An extensive shoot­ing programme (1600 km) was planned, but only 9 h of shooting on 18 April and 3 h on 14 April (a total of 98 km) were conducted owing to rough weather con-

_ _ 71*06’ N

270kg 190kg 300kg

71*02’ N

540kg 500kg 620kg

70*58’ N

1230kg 1100kg 1340kg

70*54’ N

20*05’ E 20*25’ E

w w 1 nautical mile

Figure 1. Tracklines of the geophysical survey carried out off the coast of Finnmark in 1990. Locations and catch rates of the fleets of “Husby” are shown.

63

ditions. The airgun array used had a chamber volume of 4800 in3 and the shot interval was 25 m. The by- catch of cod in the bottom trawl fishery for saithe (150-250 m bottom depth) was analysed from four trawlers fishing in the survey area before, during, and after the geophysical activity.

Results

Long-lining

Figure 1 shows the locations and catch rates of the fleets of “Husby” that were soaked on 28 and 29 January when the seismic vessel was operating. However, the exact time of setting was not noted by the skipper. Three fleets of lines were set 5-8 nautical miles from the survey area, and their mean catch rate was 1223 kg cod. The fleets set 1-5 nautical miles from the survey area caught on average 553 kg, and the fleets set within the area caught 253 kg. These catch rates are 45 and 21%, respectively, of that of the fleets set furthest away from the survey

+ 71 ‘20 ’ N

1200kg

1000kg

900kg1100kg

1000kg

. . 71*16’ N

. . 71'12’ N

. . 7 V 0 8 ’ N

. .7 V 0 4 ’ N

1200kg

71‘00 ’ N

20-00’ E

1 nautical mile

20-20’ E

Table 2. Distance from the seismic survey area, time of setting relative to sound emissions and catch rate of long-line fleets of the autoliners “Førde” and “Frøyanes” . Distance from survey area is the distance (in nautical miles) between each end of the fleet and the nearest survey trackline. Elapsed time is the time that elapsed between the ending of sound emissions and setting of fleets. Fleets 11-15 of “Frøyanes” are related to period 2 (see Table 1), the other fleets are related to period 1.

VesselDistance from

survey areaElapsed

timeCatch rate (kg/fleet)

“Førde”Fleet 1 6.5-2.6 During“ 1100Fleet 2 0.7-4.1 During3 1000Fleet 3 0.7-4.1 6 h 30 min 1200Fleet 4 0-1.2 7 h 10 min 500Fleet 5 0.8-1.8 24 h 40 min 600Fleet 6 2.0-5.3 25 h 20 min 900Fleet 7 5.7-9.7 25 h 50 min 1000Fleet 8 9.9-13.7 26 h 20 min 1200

“Frøyanes”Fleet 1 5.1-7.2 Beforeb 2500Fleet 2 6.1-8.3 Beforeb 2000Fleet 3 4.1-6.7 Beforeb 2800Fleet 4 6.7-8.4 Beforeb 2700Fleet 5 3.4-5.9 - 4 h 20 minc 1800Fleet 6 6.2-7.9 During2 1600Fleet 7 2.0-5.7 1 h 50 min 1900Fleet 8 5.0-7.2 9 h 45 min 1300Fleet 9 1.5-6.3 17 h 5 min 1300Fleet 10 3.3-6.5 24 h 35 min 2200Fleet 11 2.3-5.0 - 4 h 40 minc 1200Fleet 12 4.4-6.5 During2 1300Fleet 13 5.5-6.6 During2 1500Fleet 14 6.7-7.2 3 h 30 min 1400Fleet 15 7.3-8.2 10 h 25 min 1200

a The fleet was set during sound emissions. b The fleet was set and hauled before sound emissions

started.c Time between setting of fleet and starting of sound

emissions. The soak times for these fleets were about 9 h.

oo

o€

oO

10

8

0D A

Figure 2. Tracklines of the geophysical survey carried out off the coast of Finnmark in 1990. Locations and catch rates of the fleets of “Førde” are shown.

Figure 3. Mean by-catch of cod in shrimp trawl before (B), during (D) and more than 24 h after (A) airgun activity conducted within 5 nautical miles range from the trawler, (a) Off the coast of Finnmark, (b) In the Barents Sea east of Bear Island. Numbers of hauls are given atop each column. Note differing vertical scales.

64

B D O- 12- 24 - >4812 24 48

Figure 4. Mean by-catch of cod in bottom trawling for saithe off the coast of Møre before (B), during (D) and at different time intervals (values in hours) after airgun activity, (a) to (d) are four different fishing vessels, and (e) the average of these vessels. Number of hauls are given atop each column. Note differing vertical scales.

The catch data for “Førde” are shown in Table 2 and Figure 2. Fleets 1 and 2 were set during sound emissions, and fleets 3 and 4 about 7 h after sound emissions ended. The catch rate of fleet 4, which was set within the seismic survey area, was 45% of the mean catch rate of fleets 1-3, which were set about 1-6 nautical miles from this area. Fleets 5-8 were set about 24 h after sound emissions ended, and at increasing distances from the survey area. The catch rates for these fleets increased with increasing distance from the survey area.

The fleets of “Frøyanes” were all set about 1.5-8.5 nautical miles from the seismic survey area (Table 2). The fleets soaked prior to sound emissions caught 2000- 2800 kg cod each (mean = 2500 kg). Fleets set during sound emissions and within 24 h after emissions ended caught 1200-2200 kg (mean = 1518 kg), which indicates a catch reduction of about 40%.

The notes obtained from the “Værland” give the total catch for each day of fishing and not for each fleet. The catches for 15 to 19 January ranged from 6300 to 8250 kg

65

(mean = 7100 kg), whereas 4000 kg were caught 20 January (sound emissions started at 18.32 on 19 January, Table 1). On 21 January this vessel changed fishing

ground.

Trawling

By-catches of cod in shrimp trawls were significantly (p < 0.05) reduced during airgun activity in both areas investigated. When shooting started, cod catches de­creased from an average of 24 kg per towing hour to 5 kg per hour (79% reduction) off Finnmark (Fig. 3a). In the Bear Island area, the by-catch of cod decreased from an average of 18 kg per towing hour to 3 kg per hour (83% reduction; Fig. 3b), when the seismic vessel started shooting in the trawling area (less than 5 nautical miles from the trawler). In the Bear Island area, catches of cod increased to pre-shooting levels about one day after shooting ended.

For the trawlers operating at Storegga a threefold increase in cod catches was found when the short airgun activity sequences were conducted. The catches were significantly higher during shooting than before and at different time intervals after shooting (p < 0.05; Fig. 4a- e). When sound emission ended, the catches returned to pre-shooting levels within about 12 h. The number of hauls taken during shooting, however, was low owing to the short shooting period.

Discussion

Fish are known to hear and react to sounds (Hawkins,1986), and the results obtained in the present study demonstrate that cod respond to sounds from airgun arrays. The airguns normally used for offshore geophysi­cal exploration produce low-frequency sounds with maximum energy levels from below 10 to a few hundred Hz (Malme et al., 1986), which match well the most sensitive frequency band of cod (Chapman and Haw­kins, 1973).

Fish may react to sound in different ways. Cod have been observed to avoid an approaching sound source (vessel) with a calm swimming movement horizontally and vertically (Olsen et al. , 1982; Engås et al., 1991a, b; Ona and Godø, 1990). It has also been suggested that demersal fish such as cod dive and stay nearly immobile on the seabed when scared by noise (Dalen and Raknes, 1985). Thus, when responding to sounds from airguns, the fish may either leave the survey area or stay within the area close to the bottom. The results obtained in this study support the former explanation as both long-line and trawl catches were shown to decrease during sound emission. If the fish remained in the area close to the bottom, they should have become more available for the shrimp trawl which is designed to maintain close contact with the seabed.

However, the by-catch of cod in the trawl fishery for saithe showed a threefold increase during airgun shoot­ing. What was unusual about this case was the short duration of the shooting sequences owing to bad weather conditions. If fish start to avoid the sound source by diagonal downwards movements, the concen­tration of fish will temporarily increase at the bottom close to the survey tracklines as the fish move out of the survey area. This “plough effect” is probably of short duration, and a trawler fishing close to the tracklines when the seismic vessel starts shooting may, for a short time, make larger catches.

The present catch data were obtained from commer­cial fishing vessels that happened to be operating on fishing grounds where seismic explorations were being conducted. These data therefore do not provide an ideal basis for quantifying catch reductions caused by seismic survey operations or for investigating the spatial and temporal extent of the effects of such operations. How­ever, although uneven fish distribution may cause sig­nificant differences among catches obtained from the same fishing ground, the catch data investigated in this study, with one exception, showed the same tendency. These data therefore demonstrate significant catch re­ductions during airgun discharges, and indicate that effects lasted for 24 h and were at least 9 km in extent. However, to obtain more exact data on the effects of seismic operation on catch rates, a controlled experi­ment should be conducted.

ReferencesChapman, C. J ., and Hawkins, A. D. 1969. The importance of

sound in fish behaviour in relation to capture by trawls. FAO Fish. R ep., 62(3): 717-729.

Chapman, C. J ., and Hawkins, A. D. 1973. A field study of hearing in the cod, Gadus morhua. L. J. Comp. Physiol., 85: 147-167.

Dalen, J ., and Raknes, A. 1985. Scaring effects on fish from three-dimensional seismic surveys. Institute of Marine R e­search, Rep. No. FO 8504, 22p, Bergen, Norway.

Engås, A ., Misund, O. A ., Soldai, A . V ., Horvei, B ., and Solstad, A . 1991a. Fish behaviour and vessel noise: catch data comparisons, noise analysis and playback experiment. In The effects of trawl performance and fish behaviour on the catching efficiency of sampling trawls. Ed. by A. Engås. Dr Philos. Thesis., University of Bergen. 94 pp.

Engås, A . , Soldai, A. V ., and Øvredal, J. T. 1991b. Avoidance reactions of ultrasonic tagged cod during bottom trawling in shallow water. ICES CM 1991/B: 4 1 ,9 pp.

Hawkins, A . D. 1986. Underwater sound and fish behaviour. In The behaviour of teleost fishes, pp. 114-151. Ed. by T. J. Pitcher. Croom Helm, London.

Malme, C. I ., Smith Jr., P. W ., and Miles, P. R. 1986. Study of the effects of offshore geophysical acoustic survey operations on important commercial fisheries in California. Report No. 6125, prepared by BBN Laboratories Inc., 10 Moulton Street, Cambridge, M A 002238, for Battelle Ventura Office, 1431 Spinnaker Drive, Ventura, CA 93001, USA. 88 pp.

66

Olsen, K ., Angell, J ., Pettersen, F., and Løvik, A. 1982. Observed fish reactions to a surveying vessel with special reference to herring, cod, capelin and polar cod. FA O Fish. R ep., 100: 131-138.

Ona, E ., and Godø, O. R. 1990. Fish reactions to trawling noise: the significance for trawl sampling. Rapp. P.-v. Réun. Cons. int. Explor. M er, 189: 159-166.

Pearson, W. H ., Skalski, J. R ., and Malme, C. I. 1992. Effects of sounds from a geophysical survey device on behavior of

captive rockfish (Sebastes spp.). Can. J. Fish, aquat. Sei., 49: 1343-1356.

Skalski, J. R ., Pearson, W. H ., and Malme, C. I. 1992. Effects of sounds from a geophysical survey device on catch-per- unit-effort in a hook-and-line fishing for rockfish (Sebastes spp.). Can. J. Fish, aquat. Sei., 49: 1357-1365.

Tavolga, W. N., Popper, A. N., and Fay, R. R. (Eds.) 1981. Hearing and sound communication in fishes. Springer- Verlag, New York. 608 pp.

67